Tech info: DM500 power supplies

[digidude]

Another excellent read from gw1 @ austech

After criticising poor quality of clone DM500 power supplies for a while I figured it was time I looked at them more closely to understand how they work (or why they don't).

Those who try repairing faulty ones by replacing one or more electrolytic capacitors will sometimes find that's not enough: sometimes the replacements sizzle and burst too. And you'll find resistors overheated and burned out. What's usually happened in these cases is the regulator and/or PWM controller IC has failed, resulting in the output rising above 20V which causes a bunch of parts to burn out.

Having learned that people like Fernbay have replacements available quite cheaply I've formed the view that it's not really worth the time it takes to dismantle, repair and reglue them - certainly not if it's more than just one or two capacitors. Not everyone will agree of course, and to help those who decide to persevere I'm posting info about the couple of DM500 PSU versions I've seen so far. There may be a errors, I don't guarantee everything is 100% correct. But it gives you an idea what you're dealing with, how they work, and how to troubleshoot if you want to persevere with a repair.

If you come across a new version you might like to take some snaps and add your observations.

I've also seen cases of intermittent failure. This can be caused by hairline fractures in the PCB, or bad solder joints on heavier components (increasingly common with lead-free solder nowadays).

If you compare the schematics to reference designs from component manufacturers (eg this or this) you'll notice several short cuts taken in the DM500 supplies. For example,

• Better designs use two 470uF capacitors after the schottky diodes rather than a single 1000uF because the lower capacity parts cope better with the high ripple current.
• Better designs use 25V output capacitors instead of 16V. In fact you'll see a number of reference designs use quite small output electrolytics, eg only 100uF. This illustrates that you're better off using a smaller capacitor with a safe voltage rating than a larger one with minimal safety margin.
• Better designs specify higher rated schottky diodes to cope with transients (eg 16A instead of 3A, and higher voltage rating too)
• Better designs sometimes specify higher rated resistors (1W instead of 0.5W etc)
• Better designs specify high ripple current type for electrolytic on the output of the schottky diode
• Some designs use two schottky diodes in parallel rather than just one; some PCBs were designed for two but only one is fitted.
• Better designs specify higher voltage rating on feedback diodes to cope with transients.
• Better designs specify higher rated drive MOSFET.

Some other comments:

• There are in some cases discrepancies between suppression capacitor grades (X1, X2, Y1, Y3) marked on the PCB and the parts actually fitted. If you're not familiar with these parts it's worth taking the time to learn about them, for example here.
• Many switchmode designs take feedback from the supply side of the output filter (10uH) rather than the output side.
• The yellow kevlar (or whatever it is) tape around the transformers and heatsink is important for safety; don't remove it.
• The glue on some components is to reduce buzzing which is a common problem in switchmodes.
• You need to be careful when removing PCBs from the cases because adhesive on the rubber standoff beneath the board can be strong; you might fracture PCB tracks if you're not careful when removing it.

Be warned: an electric shock from handling a live or charged switchmode PCB can be lethal! Unless you're knowledgeable about electronics and know what you're doing you should not open, let alone attempt to modify, a switchmode power supply. Power supplies are arguably more dangerous to handle than CRT tubes because although tubes contain higher voltage the capacitors in PSUs store more charge. Additionally there is a real safety risk if you introduce a fault. That alone is good reason to purchase a replacement rather than attempt repair yourself.

 

[digidude]

First off I'll start with the DM500 supply which I'll call SUN6106C. This is a marking found on the side of the main transformer wrapped in yellow tape (you can't see the marking on these photos). The identifier isn't found on the case label unfortunately. There doesn't seem to be anything on the labels that's a reliable indicator of board version, unfortunately, which is why I've used the internal marker.

 

[digidude]

Here are the schematic and component overlay.

 

[digidude]

Next there's another marked SUN6106D, presumably a later revision. The identifier isn't evident in the photos but can be found on the transformer tape and also on the solder-side silk screen.

I expect the D version will give better performance under load than the C version, if only because of its MOSFET driver and heatsink.

 

[digidude]

Here are the D version schematic and component overlay.

I've made these reasonably high resolution in case people want to print them.

 

[digidude]

Even if you decide not to attempt repair, you can learn quite a bit by studying the circuits to figure out how they work. Component manufacturers' reference designs (such as those I linked to above) describe circuit operation in good detail so I won't go into it here. They also discuss technical constraints which designs must meet for stability, durability and standards compliance. The more you understand those issues the more you'll appreciate why careful component selection is important for reliability. A key reason clone equipment tends to be unreliable is they substitute cheaper components with lower ratings that are adequate for normal conditions but not necessarily able to cope with transients.

A good place to start learning is the datasheet for LM431 shunt regulator. When regulating properly these maintain 2.50V on their reference pin (ie the reference input is compared to 2.50V and the optocoupler is turned on/off accordingly). If you look at the DM500 power supply schematics you'll see there are 120K and 1K resistors from that pin down to ground. The total current through these is
2.5V * (1/1K + 1/120K) = 2.52mA.
As with most op-amps the current into the reference pin is negligible, which means that current must flow through the 3K9. That means the output voltage is maintained at:
2.5V + 2.52mA * 3K9 = 12.3V
So you see, even though the shunt regulator's internal reference may be high precision, the accuracy of the output voltage is only as good as those resistors. That's why 1% precision types are used.

 

[digidude]

crashing is because 1A is insufficient.

The DM500 supply needs to be 12V DC regulated with 2 amp rating minimum - preferably 2.5 amps or more. Eg Jaycar MP-3490 or Altronics M8232. Make sure you configure for positive tip!

If your receiver is a PVR with hard disk then I recommend a 3A supply.

If you're experiencing earth loop problems (eg hum or network problems) then I recommend you take the time to locate a transformer type power supply rather than a switchmode. Yes they're large, heavy and cost more but it may well solve your problem. They're becoming less common nowadays because of MEPS (low standby power) regulations but remain preferable in certain circumstances.

If you're an experimenter who's frequently unplugging and reflashing your Dreambox then a transformer type supply is a good thing to have on your workbench. You can plug/unplug it into your Dream, Phoenix or whatever without risk of blowing up RS232 ports.

Note: when hooking up RS232 cable for flashing you still need to be careful if your Dream is also connected to a switchmode-powered TV set, as tingle currents can come from them too even if your Dream power supply is transformer type. If you usually connect your Dream to a TV set while experimenting with images then your best bet is to earth your LNB lead via DMM's earthing kit, and plug that into your Dreambox before you do anything else. That guarantees your receiver is earthed and prevents any damage.

For stability and long life you need to provide your receiver and power supply with adequate ventilation (as Humax owners found out). It's particularly important for PVR receivers because of their cramped interior and extra heat generated by the hard disk. Closed cabinets with back panel and glass doors are the worst possible arrangement technically. If the wife insists on using these then you may be able to mitigate the risk to some extent by keeping space between equipment using standoffs. But really you need to avoid that situation if at all possible. Switchmode plugpacks tend to be sealed airtight which is another reason they die prematurely.